AutoML Leaderboard
| Best model | name | model_type | metric_type | metric_value | train_time | single_prediction_time |
|---|---|---|---|---|---|---|
| 1_Default_LightGBM | LightGBM | rmse | 134.753 | 17.06 | 0.0162 | |
| 2_Default_Xgboost | Xgboost | rmse | 134.579 | 6.21 | 0.0169 | |
| 3_Default_CatBoost | CatBoost | rmse | 133.496 | 17.41 | 0.0099 | |
| 4_Default_NeuralNetwork | Neural Network | rmse | 141.773 | 8.17 | 0.0389 | |
| 5_Default_RandomForest | Random Forest | rmse | 159.961 | 11.92 | 0.0838 | |
| 10_LightGBM | LightGBM | rmse | 133.831 | 12.98 | 0.0152 | |
| 6_Xgboost | Xgboost | rmse | 135.739 | 5.06 | 0.0169 | |
| 14_CatBoost | CatBoost | rmse | 133.621 | 28.04 | 0.0114 | |
| 18_RandomForest | Random Forest | rmse | 160.813 | 24.14 | 0.0914 | |
| 22_NeuralNetwork | Neural Network | rmse | 141.839 | 7.95 | 0.0314 | |
| 11_LightGBM | LightGBM | rmse | 136.251 | 7.19 | 0.0151 | |
| 7_Xgboost | Xgboost | rmse | 135.885 | 5.12 | 0.0173 | |
| 15_CatBoost | CatBoost | rmse | 134.822 | 14.75 | 0.0103 | |
| 19_RandomForest | Random Forest | rmse | 168.098 | 15.64 | 0.0808 | |
| 23_NeuralNetwork | Neural Network | rmse | 141.462 | 9.14 | 0.0329 | |
| 12_LightGBM | LightGBM | rmse | 134.518 | 13.22 | 0.015 | |
| 8_Xgboost | Xgboost | rmse | 135.005 | 5.23 | 0.017 | |
| 16_CatBoost | CatBoost | rmse | 133.739 | 16.29 | 0.0109 | |
| the best | Ensemble | Ensemble | rmse | 132.796 | 0.38 | 0.0986 |
AutoML Performance
AutoML Performance Boxplot
Features Importance (Original Scale)
Scaled Features Importance (MinMax per Model)
Spearman Correlation of Models
Summary of 4_Default_NeuralNetwork
Neural Network
- n_jobs: -1
- dense_1_size: 32
- dense_2_size: 16
- learning_rate: 0.05
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
Optimized metric
rmse
Training time
7.9 seconds
Metric details:
| Metric | Score |
|---|---|
| MAE | 108.41 |
| MSE | 20099.5 |
| RMSE | 141.773 |
| R2 | 0.738506 |
| MAPE | 0.166832 |
Learning curves
Permutation-based Importance
True vs Predicted
Predicted vs Residuals
Summary of 16_CatBoost
CatBoost
- n_jobs: -1
- learning_rate: 0.1
- depth: 7
- rsm: 1.0
- loss_function: RMSE
- eval_metric: RMSE
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
Optimized metric
rmse
Training time
15.9 seconds
Metric details:
| Metric | Score |
|---|---|
| MAE | 101.996 |
| MSE | 17886.1 |
| RMSE | 133.739 |
| R2 | 0.767302 |
| MAPE | 0.161217 |
Learning curves
Permutation-based Importance
True vs Predicted
Predicted vs Residuals
Summary of Ensemble
Ensemble structure
| Model | Weight |
|---|---|
| 10_LightGBM | 3 |
| 12_LightGBM | 1 |
| 14_CatBoost | 2 |
| 16_CatBoost | 2 |
| 2_Default_Xgboost | 1 |
| 3_Default_CatBoost | 5 |
| 7_Xgboost | 1 |
| 8_Xgboost | 1 |
Metric details:
| Metric | Score |
|---|---|
| MAE | 101.189 |
| MSE | 17634.8 |
| RMSE | 132.796 |
| R2 | 0.770571 |
| MAPE | 0.160786 |
Learning curves
True vs Predicted
Predicted vs Residuals
Summary of 11_LightGBM
LightGBM
- n_jobs: -1
- objective: regression
- num_leaves: 63
- learning_rate: 0.2
- feature_fraction: 0.5
- bagging_fraction: 1.0
- min_data_in_leaf: 10
- metric: rmse
- custom_eval_metric_name: None
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
Optimized metric
rmse
Training time
6.9 seconds
Metric details:
| Metric | Score |
|---|---|
| MAE | 104.139 |
| MSE | 18564.4 |
| RMSE | 136.251 |
| R2 | 0.758477 |
| MAPE | 0.163691 |
Learning curves
Permutation-based Importance
True vs Predicted
Predicted vs Residuals
Summary of 10_LightGBM
LightGBM
- n_jobs: -1
- objective: regression
- num_leaves: 15
- learning_rate: 0.05
- feature_fraction: 0.8
- bagging_fraction: 0.5
- min_data_in_leaf: 50
- metric: rmse
- custom_eval_metric_name: None
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
Optimized metric
rmse
Training time
12.7 seconds
Metric details:
| Metric | Score |
|---|---|
| MAE | 102.042 |
| MSE | 17910.6 |
| RMSE | 133.831 |
| R2 | 0.766983 |
| MAPE | 0.161852 |
Learning curves
Permutation-based Importance
True vs Predicted
Predicted vs Residuals
Summary of 3_Default_CatBoost
CatBoost
- n_jobs: -1
- learning_rate: 0.1
- depth: 6
- rsm: 1
- loss_function: RMSE
- eval_metric: RMSE
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
Optimized metric
rmse
Training time
17.1 seconds
Metric details:
| Metric | Score |
|---|---|
| MAE | 101.73 |
| MSE | 17821.2 |
| RMSE | 133.496 |
| R2 | 0.768146 |
| MAPE | 0.161371 |
Learning curves
Permutation-based Importance
True vs Predicted
Predicted vs Residuals
Summary of 8_Xgboost
Extreme Gradient Boosting (Xgboost)
- n_jobs: -1
- objective: reg:squarederror
- eta: 0.1
- max_depth: 7
- min_child_weight: 25
- subsample: 0.9
- colsample_bytree: 0.6
- eval_metric: rmse
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
Optimized metric
rmse
Training time
4.9 seconds
Metric details:
| Metric | Score |
|---|---|
| MAE | 102.972 |
| MSE | 18226.3 |
| RMSE | 135.005 |
| R2 | 0.762876 |
| MAPE | 0.163192 |
Learning curves
Permutation-based Importance
True vs Predicted
Predicted vs Residuals
Summary of 22_NeuralNetwork
Neural Network
- n_jobs: -1
- dense_1_size: 32
- dense_2_size: 4
- learning_rate: 0.05
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
Optimized metric
rmse
Training time
7.4 seconds
Metric details:
| Metric | Score |
|---|---|
| MAE | 108.559 |
| MSE | 20118.3 |
| RMSE | 141.839 |
| R2 | 0.738261 |
| MAPE | 0.167924 |
Learning curves
Permutation-based Importance
True vs Predicted
Predicted vs Residuals
Summary of 23_NeuralNetwork
Neural Network
- n_jobs: -1
- dense_1_size: 64
- dense_2_size: 16
- learning_rate: 0.01
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
Optimized metric
rmse
Training time
8.9 seconds
Metric details:
| Metric | Score |
|---|---|
| MAE | 108.261 |
| MSE | 20011.6 |
| RMSE | 141.462 |
| R2 | 0.739649 |
| MAPE | 0.167118 |
Learning curves
Permutation-based Importance
True vs Predicted
Predicted vs Residuals
Summary of 2_Default_Xgboost
Extreme Gradient Boosting (Xgboost)
- n_jobs: -1
- objective: reg:squarederror
- eta: 0.075
- max_depth: 6
- min_child_weight: 1
- subsample: 1.0
- colsample_bytree: 1.0
- eval_metric: rmse
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
Optimized metric
rmse
Training time
5.9 seconds
Metric details:
| Metric | Score |
|---|---|
| MAE | 102.541 |
| MSE | 18111.5 |
| RMSE | 134.579 |
| R2 | 0.764369 |
| MAPE | 0.161782 |
Learning curves
Permutation-based Importance
True vs Predicted
Predicted vs Residuals
Summary of 14_CatBoost
CatBoost
- n_jobs: -1
- learning_rate: 0.05
- depth: 8
- rsm: 0.8
- loss_function: RMSE
- eval_metric: RMSE
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
Optimized metric
rmse
Training time
27.7 seconds
Metric details:
| Metric | Score |
|---|---|
| MAE | 101.895 |
| MSE | 17854.5 |
| RMSE | 133.621 |
| R2 | 0.767713 |
| MAPE | 0.161677 |
Learning curves
Permutation-based Importance
True vs Predicted
Predicted vs Residuals
Summary of 15_CatBoost
CatBoost
- n_jobs: -1
- learning_rate: 0.1
- depth: 8
- rsm: 1.0
- loss_function: MAE
- eval_metric: RMSE
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
Optimized metric
rmse
Training time
14.4 seconds
Metric details:
| Metric | Score |
|---|---|
| MAE | 102.889 |
| MSE | 18177.1 |
| RMSE | 134.822 |
| R2 | 0.763516 |
| MAPE | 0.162411 |
Learning curves
Permutation-based Importance
True vs Predicted
Predicted vs Residuals
Summary of 6_Xgboost
Extreme Gradient Boosting (Xgboost)
- n_jobs: -1
- objective: reg:squarederror
- eta: 0.075
- max_depth: 8
- min_child_weight: 5
- subsample: 1.0
- colsample_bytree: 1.0
- eval_metric: rmse
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
Optimized metric
rmse
Training time
4.8 seconds
Metric details:
| Metric | Score |
|---|---|
| MAE | 103.439 |
| MSE | 18425.2 |
| RMSE | 135.739 |
| R2 | 0.760288 |
| MAPE | 0.162912 |
Learning curves
Permutation-based Importance
True vs Predicted
Predicted vs Residuals
Summary of 18_RandomForest
Random Forest
- n_jobs: -1
- criterion: squared_error
- max_features: 0.5
- min_samples_split: 20
- max_depth: 4
- eval_metric_name: rmse
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
Optimized metric
rmse
Training time
23.8 seconds
Metric details:
| Metric | Score |
|---|---|
| MAE | 124.897 |
| MSE | 25860.9 |
| RMSE | 160.813 |
| R2 | 0.663549 |
| MAPE | 0.183888 |
Learning curves
Permutation-based Importance
True vs Predicted
Predicted vs Residuals
Summary of 12_LightGBM
LightGBM
- n_jobs: -1
- objective: regression
- num_leaves: 63
- learning_rate: 0.05
- feature_fraction: 0.9
- bagging_fraction: 1.0
- min_data_in_leaf: 20
- metric: rmse
- custom_eval_metric_name: None
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
Optimized metric
rmse
Training time
12.8 seconds
Metric details:
| Metric | Score |
|---|---|
| MAE | 102.6 |
| MSE | 18095.2 |
| RMSE | 134.518 |
| R2 | 0.764582 |
| MAPE | 0.161761 |
Learning curves
Permutation-based Importance
True vs Predicted
Predicted vs Residuals
Summary of 5_Default_RandomForest
Random Forest
- n_jobs: -1
- criterion: squared_error
- max_features: 0.9
- min_samples_split: 30
- max_depth: 4
- eval_metric_name: rmse
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
Optimized metric
rmse
Training time
11.6 seconds
Metric details:
| Metric | Score |
|---|---|
| MAE | 123.849 |
| MSE | 25587.6 |
| RMSE | 159.961 |
| R2 | 0.667106 |
| MAPE | 0.181633 |
Learning curves
Permutation-based Importance
True vs Predicted
Predicted vs Residuals
Summary of 7_Xgboost
Extreme Gradient Boosting (Xgboost)
- n_jobs: -1
- objective: reg:squarederror
- eta: 0.1
- max_depth: 8
- min_child_weight: 1
- subsample: 1.0
- colsample_bytree: 1.0
- eval_metric: rmse
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
Optimized metric
rmse
Training time
4.8 seconds
Metric details:
| Metric | Score |
|---|---|
| MAE | 103.627 |
| MSE | 18464.7 |
| RMSE | 135.885 |
| R2 | 0.759774 |
| MAPE | 0.163594 |
Learning curves
Permutation-based Importance
True vs Predicted
Predicted vs Residuals
Summary of 1_Default_LightGBM
LightGBM
- n_jobs: -1
- objective: regression
- num_leaves: 63
- learning_rate: 0.05
- feature_fraction: 0.9
- bagging_fraction: 0.9
- min_data_in_leaf: 10
- metric: rmse
- custom_eval_metric_name: None
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
Optimized metric
rmse
Training time
16.8 seconds
Metric details:
| Metric | Score |
|---|---|
| MAE | 102.695 |
| MSE | 18158.3 |
| RMSE | 134.753 |
| R2 | 0.76376 |
| MAPE | 0.161753 |
Learning curves
Permutation-based Importance
True vs Predicted
Predicted vs Residuals
Summary of 19_RandomForest
Random Forest
- n_jobs: -1
- criterion: squared_error
- max_features: 0.7
- min_samples_split: 50
- max_depth: 3
- eval_metric_name: rmse
- explain_level: 1
Validation
- validation_type: kfold
- k_folds: 5
- shuffle: True
Optimized metric
rmse
Training time
15.3 seconds
Metric details:
| Metric | Score |
|---|---|
| MAE | 131.02 |
| MSE | 28256.8 |
| RMSE | 168.098 |
| R2 | 0.632379 |
| MAPE | 0.189525 |
Learning curves
Permutation-based Importance
True vs Predicted
Predicted vs Residuals
